utilisation of the steap protein family in a diagnostic setting may...

RESEARCH ARTICLE

Utilisation of the STEAP protein family in a

diagnostic setting may provide a more

comprehensive prognosis of prostate cancer

Stephanie E. A. BurnellID1, Samantha Spencer-Harty2, Suzie Howarth3, Owen Bodger1,

Howard Kynaston4, Claire Morgan1, Shareen H. Doak1*

1 Institute of Life Science, Swansea University Medical School, Swansea University, Singleton Park,

Swansea, Wales, United Kingdom, 2 Cellular Pathology, Abertawe Bro Morgannwg University Health Board,

Singleton Hospital, Sketty Lane, Sketty, Swansea, Wales, United Kingdom, 3 Histopathology, Abertawe Bro

Morgannwg University Health Board, Morriston Hospital, Heol Maes Eglwys, Morriston, Swansea, Wales,

United Kingdom, 4 Cardiff School of Medicine, Cardiff University, Heath Park, Cardiff, Wales, United

Kingdom

* [email protected]

Abstract

Prostate cancer is the second most common cancer diagnosed in men worldwide; however,

few patients are affected by clinically significant disease within their lifetime. Unfortunately,

the means to discriminate between patients with indolent disease and those who progress

to aggressive prostate cancer is currently unavailable, resulting in over-treatment of

patients. We therefore aimed to determine biomarkers of prostate cancer that can be used

in the clinic to aid the diagnosis and prognosis. Immunohistochemistry analysis was carried

out on prostate cancer specimens with a range of Gleason scores. Samples were stained

and analysed for intensity of the Seven Transmembrane Epithelial Antigen of the Prostate

(STEAP)-1, -2, -3, -4 and the Divalent Metal Transporter 1 (DMT1) proteins to determine

suitable biomarkers for classification of patients likely to develop aggressive prostate can-

cer. Additionally, these proteins were also analysed to determine whether any would be able

to predict future relapse using Kaplan Meier analysis. Data generated demonstrated that the

protein expression levels of STEAP2 correlated significantly with Gleason score; further-

more, STEAP4 was a significant predictor of relapse. This data indicates that STEAP2

could be potential prognostic candidate for use in combination with the current prostate can-

cer detection methods and the presence of STEAP4 could be an indicator of possible

relapse.

Introduction

Prostate cancer (PCa) is the second most common cancer worldwide, however there is cur-

rently no successful screening system [1, 2]. Men without symptoms are discouraged from

PCa screening by the US Preventive Services Task Force due to the risk of detecting slow grow-

ing cancers that will not require treatment within the patients’ lifetime [3]. As slow growing

cancers cannot be distinguished from fast growing, aggressive cancers, new prognostic

PLOS ONE | https://doi.org/10.1371/journal.pone.0220456 August 8, 2019 1 / 11

a1111111111

a1111111111

a1111111111

a1111111111

a1111111111

OPEN ACCESS

Citation: Burnell SEA, Spencer-Harty S, Howarth S,

Bodger O, Kynaston H, Morgan C, et al. (2019)

Utilisation of the STEAP protein family in a

diagnostic setting may provide a more

comprehensive prognosis of prostate cancer. PLoS

ONE 14(8): e0220456. https://doi.org/10.1371/

journal.pone.0220456

Editor: MOHAMMAD Saleem, University of

Minnesota Twin Cities, UNITED STATES

Received: February 1, 2019

Accepted: July 16, 2019

Published: August 8, 2019

Copyright: © 2019 Burnell et al. This is an open

access article distributed under the terms of the

Creative Commons Attribution License, which

permits unrestricted use, distribution, and

reproduction in any medium, provided the original

author and source are credited.

Data Availability Statement: All relevant data are

within the manuscript and its Supporting

Information files.

Funding: This work was funded from Tenovus

Cancer Care (https://tenovuscancercare.org.uk/) as

a PhD Studentship (PHD2013/L01) to SHD and

HK. The funder had no role in study design, data

collection and analysis, decision to publish, or

preparation of the manuscript.

http://orcid.org/0000-0003-3920-6180

https://doi.org/10.1371/journal.pone.0220456

http://crossmark.crossref.org/dialog/?doi=10.1371/journal.pone.0220456&domain=pdf&date_stamp=2019-08-08








http://creativecommons.org/licenses/by/4.0/

https://tenovuscancercare.org.uk/

biomarkers are required to improve patient stratification, assist with clinical management of

the disease and prevent the overtreatment of PCa patients.

The six-transmembrane epithelial antigen of the prostate (STEAP) family contains four

members (STEAP-1, -2, -3 and -4). Proteins containing this 6-transmembrane domain often

function as ion channels at cell junctions, which could be a function of the STEAP family, in

addition to their suggested metalloreductase function [4]. Due to significant sequence homol-

ogy with various metalloreductases, it has been suggested that the STEAP family may play a

role in iron and copper reduction. Indeed, STEAP2, 3 and 4 expression has been shown to

increase iron and copper uptake and promote reduction of iron and copper in vitro [5]. Before

endocytosed iron can be released from the endosome, into the cell it must be reduced and

STEAP, as a ferrireductase, can reduce Fe3+ to Fe2+ which allows its transport out of the endo-

some by the divalent metal transporter 1 (DMT1). DMT1 has been shown to be overexpressed

in colorectal cancer, consequently dysregulating iron homeostasis, resulting in tumourigenesis

[6, 7].

The STEAP protein family has been implicated in many forms of cancer due to overexpres-

sion in malignant cells when compared to their non-malignant counterparts. STEAP1 was the

first discovered and the smallest member of the STEAP family at only 339 amino acids, as it is

missing the N-terminal FNO-like domain [4]. STEAP1 localises to the plasma membrane, it

mediates the transfer of small molecules between adjacent cells in culture and is involved in

intercellular communication [8, 9]. STEAP1 is expressed in prostate epithelium and at very

low levels in a variety of other epithelial tissues including bladder tissue [10]. However,

STEAP1 is highly overexpressed in PCa along with 10 other cancers including brain, lung and

pancreatic and is associated with poor prognosis and shorter biochemical recurrence survival

[10–12]. STEAP2, also known as STAMP1, is primarily expressed in the prostate, with the

ovary being the only other significant area of expression [13]. STEAP2 has been shown to shut-

tle between the Golgi and the plasma membrane before co-localising with early endosomes

and it has been speculated to be involved in the processing and/or excretion of prostate specific

proteins such as prostate specific antigen (PSA) [14]. The expression of STEAP2 in carcinoma

and Benign Prostatic Hyperplasia (BPH) was compared by Porkka et al, with results showing

that STEAP2 expression was significantly higher in carcinoma than in BPH [13]. STEAP3, also

known as STAMP3 and TSAP6 is shown to be expressed at low levels in all tissues but gener-

ally decreased in cancer [15]. STEAP3 has been shown contain a p53-response element within

the promoter region and to be transcriptionally activated by p53 in response to stress, suggest-

ing a tumour suppressor role of STEAP3, which contrasts the other STEAP proteins [16]. Fur-

thermore, STEAP3 is strongly implicated in iron metabolism and is also reported to be

expressed at important iron metabolism sites such as the bone marrow, foetal liver and macro-

phages, in addition to partially co-localising with transferrin, transferrin receptor and DMT1

all of which are important in the transferrin cycle [17, 18]. This therefore advocates another

important role for STEAP3 within iron metabolism. STEAP 4, also known as STAMP2 and

TIARP, is primarily expressed in the heart, lung, placenta and prostate with virtually no

expression in neuronal tissue [15, 19]. STEAP4 has similar localisation to STEAP2, in addition

to localising to the vesicular-tubular structures and co-localising with early endosome antigen

1 (EEA1), suggesting a comparable secretory function to STEAP2 [19]. Silencing of STEAP4

resulted in impairment of insulin-stimulated glucose transport and STEAP4 plays a major role

in the protection of cells against metabolic deregulation and inflammatory processes both invitro and in vivo [20–22].

This information taken together has led to the hypothesis that the STEAP family, along

with DMT1, may be able to identify the extent of disease and therefore aid the prognosis of

PCa patients, resulting in the administration of more suitable treatments on a patient by

Immunohistochemical analysis of the STEAP protein family


Competing interests: The authors have declared

that no competing interests exist.


patient basis. The aim of this study was to determine the differences between protein concen-

tration and localisation of each member of the STEAP family, plus DMT1 and observe whether

any of these proteins could be utilised in a prognostic fashion in the clinic. We present data

here to indicate that STEAP2 may be utilised as a prognostic biomarker, in combination with

current methods, due to the significant correlation of STEAP2 expression and Gleason score.

Materials and methods

Tissue Microarrays (TMAs) were constructed from PCa specimens, from prostatectomy

patients, with a range of Gleason scores (n = 209 malignant and n = 47 patient-matched

healthy samples) (Wales Cancer Bank (WCB), Cardiff, UK). The Wales Cancer Bank is

licensed by the Human Tissue Authority (licence 12107) to store human tissue for research,

which has been obtained from patients providing informed consent. Wales Cancer Bank holds

ethical approval from Wales Research Ethics Committee 3 to act as a research tissue biobank

to both collect and issue biomaterials for cancer related research, where successful peer-

reviewed applications have been approved. All methods applied in this report were therefore

carried out in accordance with relevant guidelines and regulations in place for Wales Cancer

Bank related studies. Prostate samples were bordered and crossed with colorectal tumour tis-

sue to aid orientation. Hematoxylin and eosin (H&E) stained TMA slides were objectively ana-

lysed by an independent pathologist to determine that the samples were of a satisfactory

standard. Samples with insufficient staining, poor quality of stain, compromised cores (e.g.

rolled/partially missing) or lack of tumour cells in the sample were discarded from analysis.

Patient information was collected and stored in an anonymised fashion, including Gleason

score, PSA levels, age, TNM classification, death and relapse (Table 1).

The TMA slides were processed using the Benchmark XT automated staining system

(ULTRA Ventana, Singleton Hospital, Swansea, UK). Formalin fixed, paraffin embedded sec-

tions mounted on FLEX IHC slides were baked in a 60˚ oven for 1 hour. Barcode labelled

slides were placed on the Ventana machine, rinsed with reaction buffer and a liquid cover slip

was applied before Ezprep solution was used to dewax the slides at 72˚. Following antigen-

retrieval, pre-peroxidase inhibitor was applied for 4 minutes at 36˚. Antigen retrieval condi-

tions and antibody incubation conditions for each antibody utilised are described in Table 2.

OV HQ Universal linker containing secondary antibody was applied (8mins), OV- HPR mul-

timer was applied (8mins). OV DAB and H2O2 was applied (8mins) before being incubated in

copper (4mins) followed by OV AMP multimer (4-8mins). The slides were then counter-

stained with Heamatoxylin for 8 mins.

Samples were scored by two individuals, blind to patient details, based on the intensity

(0 = Negative, 1 = Weak, 2 = Moderate and 3 = strong) of the highest stain observed and the

Table 1. Patient Information.

Gleason Score STEAP1 STEAP2 STEAP3 STEAP4 DMT1 Age Range PSA Range Death/ Relapse

Normal 24 28 27 25 27 48–75 2.8–19.6 4

6 24 22 23 23 26 49–71 2.2–19.6 3

3+4 = 7 40 36 37 40 39 49–74 4.5–22.5 2

4+3 = 7 12 12 11 12 12 54–75 2.2–18 3

8 23 24 24 24 27 48–72 3.1–31.7 9

9/10 46 46 48 47 45 43–84 2.2–160 9

Total 169 168 170 171 176 43–84 2.2–160 30

Number of samples available for each biomarker, along with age range, PSA range and number of death and relapse cases for each Gleason score.

https://doi.org/10.1371/journal.pone.0220456.t001





percentage (0 = 0%, 1 = 1–25%, 2 = 25–50%, 3 = 51–75% and 4 = 75–100%) of cells stained

with this intensity, subcellular localisation was also noted. The final score was then obtained

by multiplying the intensity and percentage scores together; defined as: 0 = Negative, 1–3 =

Weak, 4–6 = Moderate and 8–12 = Strong [23, 24].

Statistical analysis

Statistical analysis was performed using SPSS software version 22. Pearson test to assess the

correlation between biomarker, Gleason, PSA and age. Differences between means (two sam-

ples) were analysed using the independent samples t-test and >two samples were analysed a

one-way ANOVA. A p-value of<0.05 was significant. Kaplan Meier plots were carried out to

determine if there was a significant difference in the time taken to relapse. Kaplan Meier analy-

sis was performed using SPSS and the long rank test was utilised to analyse this data. Due to

requirement for normally distributed data PSA values were log transformed to produce a nor-

mal distribution.

Results

The protein expression of the STEAP family, along with DMT1 were analysed in PCa samples

with varying Gleason scores using IHC. In addition to the Gleason score of the patient samples,

PSA level, age and relapse information was provided as stated in Table 1. Comparisons between

each potential biomarker and the patient information were carried out to observe the relation-

ship between current prognostic methods and identification of aggressive PCa. This evaluation

was aimed at determining whether additional prognostic information required for successful

clinical management of PCa patients could be fulfilled by evaluation the STEAP protein family.

STEAP2 protein expression significantly correlates with Gleason score

The protein concentration and localisation of each biomarker was observed and compared to

Gleason score (Figs 1 and 2). STEAP1 expression was significantly higher in the PCa speci-

mens relative to the normal prostate specimens (p<0.001) and was observed to localise to the

cytoplasm of all cells. The STEAP1 stain intensity in the normal samples was weak, the staining

increased slightly in the Gleason 6 samples and then dramatically in the Gleason 7 (3+4 and

4+3) onwards.

STEAP2 expression was significantly higher level in the PCa specimens relative to the nor-

mal prostate specimens (p<0.001) with a positive correlation between STEAP2 and Gleason

score (r = 0.308, p<0.001). STEAP2 expression was very low in the normal prostate specimens

with only minimal staining present in the cytoplasm. In contrast to this, STEAP2 was observed

Table 2. Antibodies and hybridisation conditions utilised in the IHC analysis.

Retrieval Conditions Antibody Conditions

Antibody Supplier Positive Control Buffer Time (mins) Dilution Temperature Incubation Time (mins)

STEAP1 Abcam Kidney and prostate CC1 8 1:200 RT 24

STEAP2 Abcam Prostate CC1 32 1:50 36 28

STEAP3 Abcam Pancreas and prostate CC1 32 1:80 RT 24

STEAP4 Abcam Skin and Prostate CC1 40 1:50 RT 40

DMT1 Abcam Kidney and Prostate CC1 40 1:100 36 40

The dilution, temperature and incubation time were optimised for each antibody using the positive controls specified; the final conditions are detailed here. (Cell

Conditioning 1 Buffer = CC1 and Room Temperature = RT). Corresponding images of the positive control tissue can be found in S1 Fig.






to localise to the nucleus in the PCa specimens. Stain intensity was increased dramatically in

the Gleason 6 samples. From Gleason 7 (3+4 and 4+3) the intensity of STEAP2 stain increased

in a step wise fashion to Gleason 9/10.

Fig 1. Alterations in biomarker protein expression compared against Gleason score. Panel shows the differences in

protein expression of the 4 STEAP family members plus DMT1 with regards to increasing Gleason score. (A) STEAP1

expression increases from the normal sample to the cancer samples and increases from Gleason 6 to Gleason 7. From

Gleason 7, the protein expression appears to decrease in a step-wise manner towards Gleason 9/10. (B) STEAP2

expression increases from the normal samples to the cancer samples, and then continues to increase in a step-wise

manner towards Gleason 9/10. (C) STEAP3 protein expression is consistently high in all samples observed. (D)

STEAP4 protein expression increases from the normal to cancer samples and remains high. The highest protein

expression is observed in the Gleason 7 samples. (E) DMT1 protein expression increases from the normal to cancer

samples, then remains consistently high throughout. Scale bars represent 50 μm (main image) and 20 μm (insert).

https://doi.org/10.1371/journal.pone.0220456.g001

Fig 2. Correlation between biomarker and Gleason score. Bars show Healthy patients against Gleason score

breakdown. STEAP1, 2, 4 and DMT1 protein expression was significantly increased in PCa samples compared to

healthy samples. Significant positive correlation (r = 0.308, p<0.001) was observed for STEAP2 against Gleason scores.

Significance between healthy vs cancer samples (independent samples, two-tailed t-test) denoted as �<0.05, ��<0.01,��<0.001.







STEAP3 expression was high in all samples and present at the cell membranes and the cyto-

plasm, with no difference between healthy and cancer samples and no correlation with regards

to Gleason score.

STEAP4 expression was very low in the normal prostate specimens with only minimal pro-

tein expression present in the cytoplasm. In contrast, STEAP4 localises to the lumen, with

heavy cytoplasmic staining in all cells of the PCa specimens. STEAP4 expression was signifi-

cantly higher in the PCa specimens relative to the normal prostate specimens (p<0.001),

although there was no correlation between STEAP4 and Gleason scores.

DMT1 expression was very low in the normal prostate specimens, present only in the cyto-

plasm. In contrast, DMT1 protein expression is dramatically increased and is observed to

localise to the cytoplasm of all PCa specimens. DMT1 expression was significantly higher

level in the PCa specimens when compared to the normal prostate specimens (p<0.001). No

correlation was observed between DMT1 and Gleason score, according to Pearson correlation

tests.

No correlation between LogPSA/Age and biomarker score

The two common factors in the progression of PCa are PSA value and age. As Gleason score

increases, we have observed that the PSA value and the age of the patient also increase

(r = 0.217, p = 0.011 and 0.193, p = 0.023 respectively) (Fig 3A and 3B), which indicates that

these two factors can predict the Gleason score of the patient. However, as shown in Fig 3C,

PSA value and age are also significantly positively correlated (0.231, p = 0.006), indicating that

there may be other factors influencing the increase of PSA in the body. For a biomarker to be

capable of predicting prognostic outcome, it is important that it is not affected by any other

confounding factor. In the present study, biomarker score was split into three groups: Low (0–

2), Medium (3–6) and High (8–12) and we found no correlation with either PSA value or age

for any of the biomarkers analysed (Fig 3D and 3E).

Fig 3. Correlations of PSA Value, Age, Gleason score and biomarker score. Both (A) LogPSA value and (B) Age

correlate positively with Gleason score (r = 0.217, p = 0.011 and 0.193, p = 0.023 respectively, Pearson test). (C) A

positive correlation was also observed between LogPSA and age (r = 0.231, p = 0.006, Pearson test). No correlation was

observed between STEAP1, 2, 3, 4 or DMT1 with (D) PSA value or (E) Age. Significance denoted as �<0.05, ��<0.01,��<0.001.






Chance of relapse was predicted by Gleason score, PSA value and age, while

time taken to relapse was predicted by STEAP4 expression

It is important for a prognostic biomarker to be able to determine the likely progression of the

disease. Fig 4A shows a positive correlation between increasing Gleason score and relapsing

population (r = 0.8). Fig 4B and 4C show that patients with increasing PSA levels and age were

more likely to suffer from death and relapse, however, this was not significant. Fig 4D and 4E

indicate that relapse was more likely to occur in patients with increased DMT1 expression

(p = 0.037, independent samples, two-tailed t-test). A binary logistic regression was carried out

to determine the true prognostic value of DMT1, this analysis showed the odds of relapse

increased by 20.8% per unit increase of DMT1 (p = 0.037). This increases to 22.7% per unit of

DMT1 score (p = 0.028) when controlling for Gleason, PSA and Age. This therefore indicates

that the prognostic value of DMT1 stands true when controlling for the other variable, namely

Gleason score, PSA value and age. Fig 4E illustrates the increase of relapse cases with respect to

DMT1 score, where score was split into three groups: Low (0–2), Medium (3–6) and High (8–

12). No Significant difference was observed in any STEAP biomarker expression with regards

to relapse (Fig 4D).

Kaplan Meier graphs were used to evaluate the relationship between the test biomarker

expression and time to relapse, we observed no significant association for Gleason, STEAP1,

STEAP2, STEAP3 and DMT1 scores (Fig 5A, 5B, 5C, 5D and 5F). In contrast, STEAP4 was the

only biomarker showing significant association with relapse suggesting that patients with a

high STEAP4 protein expression relapsed more quickly than those with medium or low

STEAP4 protein expressions (Fig 5E).

Discussion

Currently, Gleason scoring is the gold standard for predicting PCa outcome; however, the

Gleason score is subjective, being dependent on the pathologist’s judgement. It is therefore

important that additional biomarkers are identified that are more predictive of patients likely

to develop advanced disease. In the present study, all members of the STEAP protein family

Fig 4. Chance of relapse was increased with increasing Gleason Score, PSA level and age. (A) Gleason, (B) PSA level

and (C) Age were plotted against the death and relapse of the patient. There was a positive correlation between

increasing Gleason score and relapsing patients. (D) Relapse was significantly more likely with a higher DMT1 score

(p = 0.037, independent samples, two-tailed t-test) and (E) An increased number of relapse cases occur with a higher

DMT1 score. (High group n = 13, medium group n = 4, low group n = 0). Significance denoted as �<0.05, ��<0.01 and��<0.001.






were analysed to determine their suitability to improve the prognostic prediction of PCa. In

addition to the analysis of STEAP1-4, DMT1, was also analysed. There are significant positive

correlations between each of Gleason, PSA and age indicating they may be closely related.

These observations are supported by the literature, where correlations between these parame-

ters have previously been established [25–27]. In particular, the increase of PSA level with age

is widely documented and has resulted in the introduction of age-specific PSA reference

ranges in an attempt to increase the accuracy of PSA testing [28]. Differing PSA levels have

also been linked to race, with some studies suggesting not only age-specific, but race-specific

PSA ranges for the detection of PCa [29]. This information presents a poor case for PSA as a

diagnostic tool as there are multiple factors not related to tumourigenesis that influence PSA

level.

In the present IHC study, each biomarker protein expression score was correlated to Glea-

son score and these results indicated a significant positive correlation between STEAP2 and

Gleason score. This finding contradicts one previous study that reported no significant

Fig 5. Kaplan Meier curve as a function of Gleason score, STEAP1, 2, 3, 4 and DMT1 Score. (A) Patients with

higher Gleason scores are more likely to progress to relapse than those with lower Gleason scores, although there is no

significant difference in time taken to progress to relapse between all Gleason scores. (B) Patients with higher STEAP1

score are more likely to progress to relapse than those with medium and low scores, however, there was no significant

difference. (High group n = 17, medium group n = 10 and low group n = 7) (C) Patients with higher STEAP2 score are

more likely to progress to relapse than those with medium and low scores, however, there was no significant difference.

(High group n = 6, medium group n = 21 and low group n = 6) (D) Patients with lower STEAP3 score are less likely to

progress to relapse than those with medium and high scores, however this was not significant. (High group n = 18,

medium group n = 11 and low group n = 1) (E) Patients with higher STEAP4 score were significantly more likely to

progress to relapse earlier than those with medium or low scores (p = 0.025, long rank test). (High group n = 2,

medium group n = 18 and low group n = 10) (F) There was no difference in time taken to relapse between high and

medium DMT1 scores. (High group n = 27 and medium group n = 8).






association between STEAP2 and Gleason score [30]. However, the sample size used by the

Wang et al study (n = 67) was approximately half the samples used the present study and so

possibly lacking in power.

Since Gleason, PSA and age are all closely interlinked and elevated STEAP2 protein expres-

sion levels are significantly correlated with increasing Gleason score, it was important to deter-

mine whether the protein expression of STEAP-1-, -2, -3, -4 and DMT1 were correlated to

PSA level and age. A biomarker capable of predicting prognostic outcome that was not affected

by other factors would be beneficial. No correlation was found between any STEAP biomarker

and PSA and age respectively supporting the suitability of STEAP2 as a prognostic test for

aggressive PCa.

Although, STEAP4 was the only biomarker to be significantly associated with patient

relapse it was notable that the number of relapsing patients was very small (n = 36 out of 185

patients). These results therefore would benefit from being validated with a larger patient

cohort.

In conclusion, we suggest that STEAP1, 2, 4 and DMT1 are suitable candidates to distin-

guish patients with cancer from those that have no tumour present. When the correlation with

Gleason score was taken into consideration, STEAP2 presents as a potential prognostic bio-

marker, although on its own, it is not a solid indicator of the extent of disease. We therefore

propose STEAP2 to be used in combination with existing PCa screening methods, (e.g. PSA

testing, DRE examination and biopsy analysis). Although studies with larger cohorts of

patients are required, this biomarker is particularly promising as a predictor of prognostic out-

come due to the lack of association with confounders such as PSA levels and age.

Supporting information

S1 Fig. Immunohistochemistry positive control tissue staining for STEAP1-4 and DMT1.

Kidney and prostate tissues were used as positive controls for STEAP1 and DMT1 IHC, pros-

tate tissue was used as a positive control for STEAP2 IHC, pancreas and prostate tissues were

used as positive controls for STEAP3 IHC and skin and prostate tissues were used as positive

controls for STEAP4 IHC. Prostate tissue was from adenocarcinoma, all other tissue was

healthy. Scale bars represent 50 μm (main image) and 20 μm (insert).

(TIF)

S1 File. Raw data used to draw conclusions in this manuscript. Tables summarising the raw

data used for Table A) relationship between Gleason score and biomarker score, Table B) dif-

ference in biomarker score between tumour tissue and corresponding healthy tissue, Table C)

correlations between biomarker score, PSA and age, Table D) effect of biomarker score on the

chance of relapse, Table E) prognostic value of the biomarkers and Table F) Kaplan Meier

plots.

(DOCX)

Acknowledgments

We thank Lisa Spary for all her support and assistance in sourcing, preparation and provision

of Wales Cancer Bank prostate tissues. Tissue samples were obtained from the Wales Cancer

Bank, which is funded by the Wales Assembly Government and Cancer Research Wales.

Other investigators may have received specimens from the same subjects.

Author Contributions

Conceptualization: Shareen H. Doak.



http://www.plosone.org/article/fetchSingleRepresentation.action?uri=info:doi/10.1371/journal.pone.0220456.s001

http://www.plosone.org/article/fetchSingleRepresentation.action?uri=info:doi/10.1371/journal.pone.0220456.s002


Formal analysis: Stephanie E. A. Burnell, Owen Bodger.

Funding acquisition: Howard Kynaston, Shareen H. Doak.

Investigation: Stephanie E. A. Burnell.

Methodology: Stephanie E. A. Burnell, Samantha Spencer-Harty, Shareen H. Doak.

Software: Owen Bodger.

Supervision: Howard Kynaston, Claire Morgan, Shareen H. Doak.

Validation: Samantha Spencer-Harty, Suzie Howarth, Owen Bodger.

Visualization: Shareen H. Doak.

Writing – original draft: Stephanie E. A. Burnell.

Writing – review & editing: Samantha Spencer-Harty, Suzie Howarth, Owen Bodger, Howard

Kynaston, Claire Morgan, Shareen H. Doak.

References

1. Lloyd T, Hounsome L, Mehay A, Mee S, Verne J, Cooper A. Lifetime risk of being diagnosed with, or

dying from, prostate cancer by major ethnic group in England 2008–2010. BMC Med. 2015; 13:171.

Epub 2015/07/30. https://doi.org/10.1186/s12916-015-0405-5 PMID: 26224061; PubMed Central

PMCID: PMC4520076.

2. Pollock PA, Ludgate A, Wassersug RJ. In 2124, half of all men can count on developing prostate can-

cer. Curr Oncol. 2015; 22(1):10–2. https://doi.org/10.3747/co.22.2102 PMID: 25684983; PubMed Cen-

tral PMCID: PMC4324338.

3. Moyer VA, Force USPST. Screening for prostate cancer: U.S. Preventive Services Task Force recom-

mendation statement. Ann Intern Med. 2012; 157(2):120–34. https://doi.org/10.7326/0003-4819-157-2-

201207170-00459 PMID: 22801674.

4. Hubert RS, Vivanco I, Chen E, Rastegar S, Leong K, Mitchell SC, et al. STEAP: a prostate-specific cell-

surface antigen highly expressed in human prostate tumors. Proc Natl Acad Sci U S A. 1999; 96

(25):14523–8. https://doi.org/10.1073/pnas.96.25.14523 PMID: 10588738; PubMed Central PMCID:

PMC24469.

5. Ohgami RS, Campagna DR, McDonald A, Fleming MD. The Steap proteins are metalloreductases.

Blood. 2006; 108(4):1388–94. https://doi.org/10.1182/blood-2006-02-003681 PMID: 16609065;

PubMed Central PMCID: PMC1785011.

6. Xue X, Taylor M, Anderson E, Hao C, Qu A, Greenson JK, et al. Hypoxia-inducible factor-2α activation

promotes colorectal cancer progression by dysregulating iron homeostasis. Cancer Res. 2012; 72

(9):2285–93. Epub 2012/03/14. https://doi.org/10.1158/0008-5472.CAN-11-3836 PMID: 22419665;

PubMed Central PMCID: PMC3342485.

7. Brookes MJ, Hughes S, Turner FE, Reynolds G, Sharma N, Ismail T, et al. Modulation of iron transport

proteins in human colorectal carcinogenesis. Gut. 2006; 55(10):1449–60. Epub 2006/04/26. https://doi.

org/10.1136/gut.2006.094060 PMID: 16641131; PubMed Central PMCID: PMC1856421.

8. Challita-Eid PM, Morrison K, Etessami S, An Z, Morrison KJ, Perez-Villar JJ, et al. Monoclonal antibod-

ies to six-transmembrane epithelial antigen of the prostate-1 inhibit intercellular communication in vitro

and growth of human tumor xenografts in vivo. Cancer Res. 2007; 67(12):5798–805. https://doi.org/10.

1158/0008-5472.CAN-06-3849 PMID: 17575147.

9. Yamamoto T, Tamura Y, Kobayashi JI, Kamiguchi K, Hirohashi Y, Miyazaki A, et al. Six-transmem-

brane epithelial antigen of the prostate-1 plays a role for in vivo tumor growth via intercellular communi-

cation. Exp Cell Res. 2013; 319(17):2617–26. https://doi.org/10.1016/j.yexcr.2013.07.025 PMID:

23916873.

10. Moreaux J, Kassambara A, Hose D, Klein B. STEAP1 is overexpressed in cancers: a promising thera-

peutic target. Biochem Biophys Res Commun. 2012; 429(3–4):148–55. https://doi.org/10.1016/j.bbrc.

2012.10.123 PMID: 23142226.

11. Gomes IM, Arinto P, Lopes C, Santos CR, Maia CJ. STEAP1 is overexpressed in prostate cancer and

prostatic intraepithelial neoplasia lesions, and it is positively associated with Gleason score. Urol Oncol.

2013. https://doi.org/10.1016/j.urolonc.2013.08.028 PMID: 24239460.



https://doi.org/10.1186/s12916-015-0405-5

http://www.ncbi.nlm.nih.gov/pubmed/26224061

https://doi.org/10.3747/co.22.2102


https://doi.org/10.7326/0003-4819-157-2-201207170-00459

https://doi.org/10.7326/0003-4819-157-2-201207170-00459


https://doi.org/10.1073/pnas.96.25.14523


https://doi.org/10.1182/blood-2006-02-003681


https://doi.org/10.1158/0008-5472.CAN-11-3836


https://doi.org/10.1136/gut.2006.094060

https://doi.org/10.1136/gut.2006.094060


https://doi.org/10.1158/0008-5472.CAN-06-3849

https://doi.org/10.1158/0008-5472.CAN-06-3849


https://doi.org/10.1016/j.yexcr.2013.07.025


https://doi.org/10.1016/j.bbrc.2012.10.123

https://doi.org/10.1016/j.bbrc.2012.10.123


https://doi.org/10.1016/j.urolonc.2013.08.028



12. Ihlaseh-Catalano SM, Drigo SA, de Jesus CM, Domingues MA, Trindade Filho JC, de Camargo JL,

et al. STEAP1 protein overexpression is an independent marker for biochemical recurrence in prostate

carcinoma. Histopathology. 2013; 63(5):678–85. https://doi.org/10.1111/his.12226 PMID: 24025158.

13. Porkka KP, Helenius MA, Visakorpi T. Cloning and characterization of a novel six-transmembrane pro-

tein STEAP2, expressed in normal and malignant prostate. Lab Invest. 2002; 82(11):1573–82. PMID:

12429817.

14. Korkmaz KS, Elbi C, Korkmaz CG, Loda M, Hager GL, Saatcioglu F. Molecular cloning and characteri-

zation of STAMP1, a highly prostate-specific six transmembrane protein that is overexpressed in pros-

tate cancer. J Biol Chem. 2002; 277(39):36689–96. https://doi.org/10.1074/jbc.M202414200 PMID:

12095985.

15. Grunewald TG, Bach H, Cossarizza A, Matsumoto I. The STEAP protein family: versatile oxidoreduc-

tases and targets for cancer immunotherapy with overlapping and distinct cellular functions. Biol Cell.

2012; 104(11):641–57. https://doi.org/10.1111/boc.201200027 PMID: 22804687.

16. Passer BJ, Nancy-Portebois V, Amzallag N, Prieur S, Cans C, Roborel de Climens A, et al. The p53-

inducible TSAP6 gene product regulates apoptosis and the cell cycle and interacts with Nix and the

Myt1 kinase. Proc Natl Acad Sci U S A. 2003; 100(5):2284–9. https://doi.org/10.1073/pnas.

0530298100 PMID: 12606722; PubMed Central PMCID: PMC151332.

17. Zhang F, Tao Y, Zhang Z, Guo X, An P, Shen Y, et al. Metalloreductase Steap3 coordinates the regula-

tion of iron homeostasis and inflammatory responses. Haematologica. 2012; 97(12):1826–35. https://

doi.org/10.3324/haematol.2012.063974 PMID: 22689674; PubMed Central PMCID: PMC3590089.

18. Graham RM, Chua AC, Herbison CE, Olynyk JK, Trinder D. Liver iron transport. World J Gastroenterol.

2007; 13(35):4725–36. https://doi.org/10.3748/wjg.v13.i35.4725 PMID: 17729394.

19. Korkmaz CG, Korkmaz KS, Kurys P, Elbi C, Wang L, Klokk TI, et al. Molecular cloning and characteriza-

tion of STAMP2, an androgen-regulated six transmembrane protein that is overexpressed in prostate

cancer. Oncogene. 2005; 24(31):4934–45. https://doi.org/10.1038/sj.onc.1208677 PMID: 15897894.

20. Cheng R, Qiu J, Zhou XY, Chen XH, Zhu C, Qin DN, et al. Knockdown of STEAP4 inhibits insulin-stimulated

glucose transport and GLUT4 translocation via attenuated phosphorylation of Akt, independent of the effects

of EEA1. Mol Med Rep. 2011; 4(3):519–23. https://doi.org/10.3892/mmr.2011.443 PMID: 21468601.

21. Konstantopoulos N, Foletta VC, Segal DH, Shields KA, Sanigorski A, Windmill K, et al. A gene expres-

sion signature for insulin resistance. Physiol Genomics. 2011; 43(3):110–20. https://doi.org/10.1152/

physiolgenomics.00115.2010 PMID: 21081660.

22. Wellen KE, Fucho R, Gregor MF, Furuhashi M, Morgan C, Lindstad T, et al. Coordinated regulation of

nutrient and inflammatory responses by STAMP2 is essential for metabolic homeostasis. Cell. 2007;

129(3):537–48. https://doi.org/10.1016/j.cell.2007.02.049 PMID: 17482547; PubMed Central PMCID:

PMC2408881.

23. Fedchenko N, Reifenrath J. Different approaches for interpretation and reporting of immunohistochem-

istry analysis results in the bone tissue—a review. Diagn Pathol. 2014; 9:221. Epub 2014/11/29. https://

doi.org/10.1186/s13000-014-0221-9 PMID: 25432701; PubMed Central PMCID: PMC4260254.

24. Rizzardi AE, Zhang X, Vogel RI, Kolb S, Geybels MS, Leung YK, et al. Quantitative comparison and

reproducibility of pathologist scoring and digital image analysis of estrogen receptor β2 immunohis-

tochemistry in prostate cancer. Diagn Pathol. 2016; 11(1):63. Epub 2016/07/11. https://doi.org/10.

1186/s13000-016-0511-5 PMID: 27401406; PubMed Central PMCID: PMC4940862.

25. Lojanapiwat B, Anutrakulchai W, Chongruksut W, Udomphot C. Correlation and diagnostic perfor-

mance of the prostate-specific antigen level with the diagnosis, aggressiveness, and bone metastasis

of prostate cancer in clinical practice. Prostate Int. 2014; 2(3):133–9. https://doi.org/10.12954/PI.14054

PMID: 25325025; PubMed Central PMCID: PMC4186957.

26. Pepe P, Pennisi M. Gleason score stratification according to age at diagnosis in 1028 men. Contemp

Oncol (Pozn). 2015; 19(6):471–3. https://doi.org/10.5114/wo.2015.56654 PMID: 26843845; PubMed

Central PMCID: PMC4731454.

27. Babaian RJ, Miyashita H, Evans RB, Ramirez EI. The distribution of prostate specific antigen in men

without clinical or pathological evidence of prostate cancer: relationship to gland volume and age. J

Urol. 1992; 147(3 Pt 2):837–40. https://doi.org/10.1016/s0022-5347(17)37400-1 PMID: 1371558.

28. DeAntoni EP. Age-specific reference ranges for PSA in the detection of prostate cancer. Oncology (Will-

iston Park). 1997; 11(4):475–82, 85; discussion 85–6, 89. PMID: 9130271.

29. Gupta A, Gupta D, Raizada A, Gupta NP, Yadav R, Vinayak K, et al. A hospital based study on refer-

ence range of serum prostate specific antigen levels. Indian J Med Res. 2014; 140(4):507–12. PMID:

25488444; PubMed Central PMCID: PMC4277136.

30. Wang L, Jin Y, Arnoldussen YJ, Jonson I, Qu S, Maelandsmo GM, et al. STAMP1 is both a proliferative

and an antiapoptotic factor in prostate cancer. Cancer Res. 2010; 70(14):5818–28. https://doi.org/10.

1158/0008-5472.CAN-09-4697 PMID: 20587517.



https://doi.org/10.1111/his.12226



https://doi.org/10.1074/jbc.M202414200


https://doi.org/10.1111/boc.201200027


https://doi.org/10.1073/pnas.0530298100

https://doi.org/10.1073/pnas.0530298100


https://doi.org/10.3324/haematol.2012.063974

https://doi.org/10.3324/haematol.2012.063974


https://doi.org/10.3748/wjg.v13.i35.4725


https://doi.org/10.1038/sj.onc.1208677


https://doi.org/10.3892/mmr.2011.443


https://doi.org/10.1152/physiolgenomics.00115.2010

https://doi.org/10.1152/physiolgenomics.00115.2010


https://doi.org/10.1016/j.cell.2007.02.049


https://doi.org/10.1186/s13000-014-0221-9

https://doi.org/10.1186/s13000-014-0221-9


https://doi.org/10.1186/s13000-016-0511-5

https://doi.org/10.1186/s13000-016-0511-5


https://doi.org/10.12954/PI.14054


https://doi.org/10.5114/wo.2015.56654


https://doi.org/10.1016/s0022-5347(17)37400-1




https://doi.org/10.1158/0008-5472.CAN-09-4697

https://doi.org/10.1158/0008-5472.CAN-09-4697



utilisation of the steap protein family in a diagnostic setting may...

Documents